EasyDeepLearn

Categorical vs Gaussian vs beta policies — which and when?

medium

Answer

  • Categorical: discrete actions, softmax over logits.
  • Gaussian: continuous, N(μθ(s),  σθ(s)2)N({\mu}{\theta}(s), \;{\sigma}{\theta}(s)^{2}). σ can be state-dependent or state-independent parameter.
  • Beta: bounded continuous [a, b] → avoids clipping bias of unbounded Gaussian.
  • Tanh-Gaussian (squashed): Gaussian pass through tanh → bounded.
  • Modern default: tanh-Gaussian (SAC uses it).
  • Beta useful for asymmetric action bounds.
Check yourself — multiple choice
  • Only Gaussian
  • Categorical (discrete), Gaussian (continuous), Beta (bounded interval, no clip bias), Tanh-Gaussian (SAC default) — squashed to bounded; Beta for asymmetric bounds
  • Random
  • Not real

Policy: Categorical / Gaussian / Beta / Tanh-Gaussian; SAC uses Tanh-Gaussian.

#policy-methods

Practise Reinforcement Learning

214 interview questions in this topic.

Related questions